Induced motion and certain other illusions concerning the path of perceived motion are examples of that class of phenomena in which perception is determined by the relation of one stimulus object to another. The proposed research will investigate these phenomena as part of the long-range goal to explain perception in terms of an inference or problem-solving process. The guiding hypotheses are: (1) In induced motion the physically stationary object appears to move egocentrically because the surrounding moving object becomes the frame of reference and consequently is misperceived asegocentrically stationary or as moving less than it is. To eliminate conflict with this perceptual interpretation, the eyes (which are stationary in fixating the stationary object) are felt to be tracking a moving object. (2) In many illusions concerning the perceived path of an object moving in the vicinity of a moving frame of reference, the display is perceived to be hierarchically organized: the object moves relative to the frame and the frame moves relative to the observer. The object is seen as belonging to the moving frame of reference and as partaking of its motion. The two components of perceived motion of the object together fully account for its displacement in space. Thus the net perception of egocentric change is veridical, there is no conflict between different sources of information and, therefore, this kind of hierarchical effect is robust in comparison to induced movement. Experiments are proposed to test these hypotheses and to explore similarities and differences between the two kinds of motion illusion in order to increase our understanding of the perceptual processes that underlie normal vision. The problem is one of fundamental importance in psychology and in many areas in the field of mental health.